The Meme and the Operant

Sandy Hobbs and David Cornwell | Letters to Ambrose Merton # 27, 2002

This is an edited version of a paper read at the Association for Behavior Analysis International Conference, Venice, November 2001. Although written primarily for psychologists, we hope it will be of interest to a wider audience. It contrasts a behaviourist approach to contemporary legend (derived from B.F. Skinner’s concept of “operant conditioning”) with an approach derived from Richard Dawkins’s concept of the “meme”.

The concept of a “meme” was first proposed by Richard Dawkins, in his book The selfish gene (1976). This remarkable volume, aimed simultaneously at expert, student and layman, has been the subject of lively debate. Put briefly, Dawkins proposed a new emphasis in Darwinian thinking. Whereas Darwin had focused on the survival and evolution of species, Dawkins argued that to focus on the gene was now more powerful analytically. To this relatively prosaic proposal, Dawkins added a metaphor, taking the term “selfish”, normally applied to human behavior, and attached it to the gene as a way of highlighting his point. The fact that “selfishness” is not itself a particularly precise concept may be noted.

Returning to the more prosaic side of Dawkins’s case, the next significant feature is that he treats the gene as a “replicator”. Another example of a replicator which he suggested is something for which he felt obliged to invent a new term, the “meme”. “Meme” is suggested as a cultural equivalent to the gene. Although described rather casually in the thirteen page final chapter of Dawkins’s book, the term has met with some success. Sampling the World Wide Web on 29 August 1998, Dawkins found 5042 mentions of the adjectival form “memetic” (Dawkins, 1999). A number of books have been devoted to the meme, most notably Susan Blackmore’s The meme machine (1999), which includes a Foreword by Dawkins in which he writes (1999, p xvi):

Any theory deserves to be given its best shot, and that is what Susan Blackmore has given the theory of the meme.
Dawkins is not alone in praising the book. Another biologist, Matt Ridley, writing on Darwin in a series entitled “Giants Refreshed” for the Times Literary Supplement, presents Dawkins and Blackmore as the latest key stages in the development of Darwinian thought:

The theory of the selfish gene has swept all before it in modern evolutionary theory… But ironically, an even more unsettling theory is beginning to replace this approach: memes… Susan Blackmore has demonstrated that just as our bodies are the victims and vehicles of genes, so our minds may be victims and vehicles for ideas or memes… All this is a long way from Darwin himself, but it is his true intellectual legacy… (1999, p 13)

Note the terminology here; to the “selfishness” of the gene we now find added the idea that human beings are “victims” of memes.

Are we then in the presence of a major development in intellectual history? Perhaps, but lest we get carried away, we should note that in a subsequent issue of the same periodical, another biologist, Matthew Cobb, referred to the idea that the fittest memes survive as “tautological” and suggested it is merely a “silly dinner-party idea” (2000, p 17). Despite the enthusiastic acceptance of the concept of the meme by some writers, approval of the meme is by no means universal.

Given that there are innumerable non-behavioral terms in use which are applied to human behavior, it may be necessary at this stage to offer some justification for casting a behavioural eye over this particular example. There are two which are fairly easy to explain. First, Dawkins, who formulated the term, has written sympathetically about Skinner’s concept of selection by consequences (Dawkins, 1988), although as far as we are aware without attempting to relate the concept of the meme to that of the operant. Secondly, in her book on the meme, Blackmore, although a psychologist of a non-behavioral bent, nevertheless argues that much human behavior emerges according to the laws of operant conditioning. She acknowledges that much of what humans do is built and shaped by consequences. We tend to repeat actions which are followed by rewarding outcomes.

However, the third reason is a little more complex. It concerns a set of phenomena known usually as “contemporary legends” or “urban legends”. For several years, we have been attempting to bring a little behaviorist precision to a scholarly field where, in our view, vagueness has played too big a part (see, for example, Hobbs, 1987, Hobbs, 1989, Hobbs and Cornwell, 1991). One argument we have put forward is that the very concept “contemporary legend” should be questioned. Analysis might show that the phenomena currently so labelled might be better seen as examples of a broader, more carefully defined category. One of the leading legend scholars has suggested that this point of view which we have advocated has similarities to the suggestions of those who employ the concept of a “meme” (Brunvand, 2001). This gave us pause for thought and further encouraged us to look more closely at the meme.

But to which phenomena is the term “meme” applied? Dawkins (1976) , initially lists as examples, “tunes, ideas, catch-phrases, clothes fashions, ways of making pots or of building arches” (p 206) and later includes “the idea of God” (p 207). Blackmore deals with a wide range of topics including the origins of language, religion, the concept of an inner self and altruism. However, in explaining the concept of a meme, conveniently enough for the purposes of this paper, she starts with a contemporary legend:

Have you heard the one about the poodle in the microwave? An American lady, so the story goes, used to wash her poodle and dry it in the oven. When she acquired a brand new microwave oven, she did the same thing. Bringing the poor dog to a painful and untimely death. Then she sued the manufacturers for not providing a warning “Do not dry your poodle in this oven” – and won! (Blackmore, 1999, p 14)

Like Dawkins, Blackmore argues that a meme may be regarded as the cultural equivalent of the biological concept, the gene. The gene is a replicator and has three defining characteristics, variation, selection and retention. A meme is claimed to have those same characteristics and, at least at a superficial level, the case can be made that a story such as the one just cited has the characteristics of a replicator and hence may justifiably be termed a “meme”. These stories are told and retold, they change in the telling, and, since not all stories are passed on, those which are may thus been seen as having been selected.

However, it must be noted Blackmore deals with this contemporary legend (and indeed with other examples of memes) in a rather casual fashion. She does not refer to any specific texts collected by legend scholars and does not refer to any of the scholarly literature on such stories (see, for example, Bennett and Smith, 1993, Brunvand, 2001). In Blackmore’s defence, it might be argued that this scholarly literature has not so far presented its data in a form which would encourage her to use it. To justify the claim that contemporary legends show the characteristics of retention, variability and selection, one would need evidence which allowed us to compare different versions of a story over time. At present that is difficult to do. Difficult but not impossible.

As an illustration of the possibilities, we shall take another contemporary legend, usually referred to as The Boyfriend’s Death.

Figure 1

There were two people parked along this dark road. They were both drinking beer and after several beers the boy had to take a leak. He left the car and disappeared into the woods to position himself behind a tree. After several minutes the girl had become worried because the boy hadn’t returned. From the roof she heard a light tapping sound. She got out to call him but noticed something in the dark hanging above the car from a large tree. It was her boyfriend and he had been hung by the neck.

Text A1, collected in Kansas, 1964 (Barnes, 1966)

This single text tells us little and does not itself throw much light on whether or not we are in the presence of a meme, so we shall compare it with that in Figure 2, a text collected over twenty years later in Scotland.

Figure 2

The story was that there was a young man who had come from Aviemore who had taken his wife on honeymoon to Australia. And they were out driving in the car and the car had run out of petrol. So he was going off to get some petrol, telling her not to get out of the car. He goes off and time passes and she falls asleep. She wakes up in the darkness with a banging noise on the roof of the car and the police car sitting up ahead. And the police approach her, take her out of the car, and take her up to the police car and tell her not to look back. And she looks back and she finds an aborigine man sitting on top of the car with a pole and her husband’s head on top of the pole.

Text B3, collected by Sandy Hobbs, Paisley, 1988.

Even a casual inspection shows that there are differences between the two. The later text is about fifty per cent longer (143 words compared to 93). It contains detailed references to places, Aviemore and Australia, and to people, an aboriginal man and the police. Nevertheless the two texts contain enough in common to justify treating this as a single story: a man and a woman in a car, the man leaves the car, the man is later found dead. However, calling these texts “the same” story implies not just that there are elements in common but also that there are chains of retellings, in the course of which variations occur.

How plausible is such a claim? The main justification lies in the fact that stories with these three basic elements have been collected in different locations of the United States and the British Isles over the intervening years. Table 1 brings together information on 14 different texts (including those laid out in Figures 1 and 2). However, the manner of presentation differs from that usually adopted in contemporary legend scholarship. The common features and certain other recurring elements have been abstracted from the texts published and presented in tabular form in as near to chronological order as the information published about these texts allows us to determine.

Although these represent only a tiny fraction of the many times the story has probably been told, the information provided from this sample could actually be used by Blackmore in support of at least part of her case. The story has been retained over time and replicated. The story shows variations. The question of selection is less straightforward. We have no evidence here of this story being “selected” rather than any others. However, given that some elements not in the earliest texts become almost fixed in later versions, the story may be seen as being changed in ways which faciliate its survival.

Table 1 (a)

TEXT(b)
A1
A2
A3
A4
A5
B1
A6
A7
A8
A9
A10
B2
A11
B3

Man and woman in car
X
X
X
X
X
X
X
X
X
X
X
X
X
X

On a date
X
X
X
X
X
 
X
X
X
X
 
 
 
 

Warning heard
 
 
 
 
 
 
 
 
 
X
 
X
 
 

Out of gas/petrol
 
 
X
 
 
 
 
X
X
X
X
X
X
X

Man leaves car
X
X
X
X
X
X
X
X
X
X
X
X
X
X

Woman hears noises
X
X
X
X
X
 
X
X
X
X
 
 
 
X

Police arrive
 
 
X
X
X
X
 
X
X
X
 
X
X
X

“Don’t look back”
 
 
X
X
X
X
 
 
X
X
 
X
 
X

Man murdered
X
X
X
X
X
X
X
X
X
X
X
X
X
X

Man found (c)
H
H
UD
U
U
D
H
U
H
H
D
D
D
D

Killer Seen
 
 
 
 
 
X
 
 
 
 
X
X
X
X

Collected
’64
’64
’68
’68
’71
’76
’79
’79
’79
’79
’79
’80
’82
’88

(a) X indicates present
(b) A USA / B British Isles
(c) H Hanged U Hanged upsidedown D Decapitated

The state of the dead man becomes more gruesome, initially hanging, then hanging upsidedown, then decapitated. The sight of the killer adds to the horror. The woman’s discovery of her companion’s fate is made more dramatic, first, by the arrival of the police and, secondly, by their warning “Don’t look back”. If these changes make the story more effective, then these variations may indeed facilitate the story’s survival.

Thus it might be suggested that if Blackmore were to look more carefully than she has so far at evidence collected by legend scholars she would not find anything likely to discomfort her. At this point, however, it is perhaps necessary to recall that the title of this paper is “The meme and the operant”. This means that another question must be raised. Is the evidence summarized in Table 1 sufficient to justify the need for the concept of the meme, as opposed to those proposed by writers adopting the perspective of behavior analysis.

Writing specifically on contemporary legend from an explicitly behaviorist perspective is fairly limited. We offered a simple behavior analysis model about a decade ago (Hobbs and Cornwell, 1991) and a social contingency analysis by Guerin and Miyazaki is currently in press. We would suggest that the evidence in Table 1 is entirely compatible with such analyses. For example, if, as Guerin and Miyazaki propose, the primary functions of such stories are to hold the attention of the audience and to entertain them, then the changes noted may be seen as faciltitating these processes. The arrival of the police and the words of the warning hold the audience in suspense as they await the disclosure to the man’s fate. It has been demonstrated by Slotkin (1988) that the same person may tell the same story quite differently on different occasions. Amongst other outcomes, this gives the individual the opportunity to learn which variations are most effective.

Our argument then is that the data collected to throw light on the plausibility of the meme, however limited, may also be interpreted as demonstrating the plausibility of an operant analysis. Is this outcome likely to be restricted to contemporary legends as opposed to other forms of cultural transmission. We think not. Consider an analysis present by Dawkins in the second (1989) edition of The Selfish Gene.

Dawkins is quite explicit in claiming the physical character of memes, writing that “If memes in brains are analogous to genes they must be self-replicating brain structures, actual patterns of neuronal wiring-up that reconstitute themselves in one brain after another” (Dawkins, 1989, p 323). He goes on to distinguish between memes and their “phenotypic” effects. However, when discussing the supposed processes whereby memes function, he is forced to deal in these “phenotypes”. In one example (Dawkins, 1989, p 328) he demonstrates that he and E. O. Wilson were independently responsible for a “mutant meme”, when they incorrectly cited a paper actually called “The genetic evolution of social behaviour” as “The genetic theory of social behaviour”. This mutated title was subsequently adopted by other authors. However, the use of this quasi-biological terminology is surely unnecessary. Dawkins himself suggests that the “mutation” was not random, but was influenced by the title of a famous work in their field “The genetic theory of natural selection”. This quite plausible suggestion is surely quite easily accommodated within an analysis which treats these titles as topographically similar operants. Dawkins’s suggestion of an underlying neuronal wiring adds nothing to our understanding of these events.

The memetic and the operant approaches to the evolution of human culture have some features in common. This is clear in the following passage from Baum’s Understanding behaviorism:

A group’s culture consists of learned behavior shared by members of the group, acquired as a result of membership of the group, and transmitted from one group member to another. Evolution of culture occurs in a manner parallel to shaping of operant behavior and genetic evolution – variation coupled with selective transmission. The unit of selection – the things that vary and are selectively transmitted – are replicators. A replicator is any entity capable of producing copies of itself. A good replicator possesses longevity, fecundity, copying fidelity and efficacy. (Baum, 1994, p 231)

The key differences between the approaches is surely a matter of parsimony. To the behavior analyst the reference to underlying neural structures is unnecessary as it adds nothing to our understanding. However, given that both parties deal for the most part in generalities in discussing culture and have in the past been unable to cite evidence beyond the anecdotal level, it is difficult to imagine either side succeeding in demonstrating the superioity of their model. However, we do not suggest that we should conclude from this that dialogue is pointless.

We challenge the advocates of the “meme” on two fronts. First, as Baum has done, we should question what the reference to a supposed neural underpinning is adding to our understanding of cultural replicators. Secondly, we invite them to produce or consider data on transmission so that the relative strengths of memetic and operant analyses may be tested. It is in this latter respect that revised ways of examining contemporary legends may play a useful part.

It is probably necessary to end by making clear what we are saying about the data in Table 1. First, although it is “abstracted” from the various texts, it is nevertheless based on verbal behavior. Secondly, although the data contained in Table 1 is obviously limited, there are substantial possibilities of expanding it, by adding information on other tellings of the same story, by undertaking similar analyses of other stories and even for generating analogous data in controlled experimentation.

References

Barnes, D. R. (1966) Some functional horror stories on the Kansas University campus, Southern Folklore Quarterly, 30, 305-312.

Baum, W. M. (1994) Understanding behaviorism: Science, behavior and culture. New York: HarperCollins.

Bennett, G. and Smith, P. (eds.) (1993) Contemporary legend: A folklore bibliography. New York: Garland.

Blackmore, S. (1999) The meme machine. London: Oxford University Press.

Brunvand, J. H. (2001) Encyclopaedia of urban legends. Denver CO: ABC-Clio.

Cobb, M. (2000, 18 February) Darwin and human behaviour, Times Literary Supplement, 17.

Dawkins, R. (1976) The selfish gene. London: Oxford University Press.

Dawkins, R. (1988) Replicators, consequences, and displacement activities, pp 33-35 in Catania, A. C. and Harnard, S. (eds.) The selection of behavior: The operant behaviorism of B. F. Skinner: Comments and consequences. Cambridge: Cambridge University Press.

Dawkins, R. (1989) The selfish gene. (2nd Ed.) London: Oxford University Press.

Dawkins, R. (1999) Foreword, pp vii-xvii in Blackmore, S. The meme machine. London: Oxford University Press.

Guerin, B. and Miyazaki, Y. (forthcoming) Rumors, gossip, and urban legends: A social contingency theory, Summa Psicologica.

Hobbs, S. (1987) The social psychology of a “good” story, pp. 133-148 in Bennett, G., Smith, P. and Widdowson, J. D. A. (eds.) Perspectives on contemporary legend Volume II. Sheffield: Sheffield Academic Press.

Hobbs, S. (1989) Enough to constitute a legend?, pp 55-75 in Bennett, G. and Smith, P. (eds.) The questing beast: Perspectives on contemporary legend Volume IV. Sheffield: Sheffield Academic Press.

Hobbs, S, and Cornwell, D. (1991) A behavior analysis model of contemporary legend, Contemporary Legend, 1, 93-106.

Ridley, M. (2000, 28 January) From the bottom up, Times Literary Supplement, 13-14.
Slotkin, E. (1988) Legend genre as a function of audience, pp 89-111 in Bennett, G. and Smith, P. (eds.) Monsters with iron teeth: Perspectives on contemporary legend III. Sheffield: Sheffield Academic Press.

Note: The correct citation for the paper noted above as “Guerin and Miyazaki (forthcoming)” is:
Guerin, B and Miyazaki, Y. (2003) Rumores, chisme y leyenda urbanas: una teoria de la contingencia social, Revista latinoamericana de Psicologia, 35, 257-272.